Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L27/00—Modulated-carrier systems
  • H04L27/26—Systems using multi-frequency codes
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L27/00—Modulated-carrier systems
  • H04L27/26—Systems using multi-frequency codes
  • H04L27/2601—Multicarrier modulation systems
  • H04L27/2647—Arrangements specific to the receiver only
  • H04L27/2655—Synchronisation arrangements
  • H04L27/2668—Details of algorithms
  • H04L27/2673—Details of algorithms characterised by synchronisation parameters
  • H04L27/2676—Blind, i.e. without using known symbols
  • H04L27/2678—Blind, i.e. without using known symbols using cyclostationarities, e.g. cyclic prefix or postfix
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03006—Arrangements for removing intersymbol interference
  • H04L25/03159—Arrangements for removing intersymbol interference operating in the frequency domain
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L27/00—Modulated-carrier systems
  • H04L27/26—Systems using multi-frequency codes
  • H04L27/2601—Multicarrier modulation systems
  • H04L27/2602—Signal structure
  • H04L27/2605—Symbol extensions, e.g. Zero Tail, Unique Word [UW]
  • H04L27/2607—Cyclic extensions
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L27/00—Modulated-carrier systems
  • H04L27/26—Systems using multi-frequency codes
  • H04L27/2601—Multicarrier modulation systems
  • H04L27/2647—Arrangements specific to the receiver only
  • H04L27/2655—Synchronisation arrangements
  • H04L27/2656—Frame synchronisation, e.g. packet synchronisation, time division duplex [TDD] switching point detection or subframe synchronisation
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L1/00—Arrangements for detecting or preventing errors in the information received
  • H04L1/004—Arrangements for detecting or preventing errors in the information received by using forward error control
  • H04L1/0056—Systems characterized by the type of code used
  • H04L1/0071—Use of interleaving
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03006—Arrangements for removing intersymbol interference
  • H04L2025/0335—Arrangements for removing intersymbol interference characterised by the type of transmission
  • H04L2025/03375—Passband transmission
  • H04L2025/03414—Multicarrier
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03006—Arrangements for removing intersymbol interference
  • H04L2025/03433—Arrangements for removing intersymbol interference characterised by equaliser structure
  • H04L2025/03439—Fixed structures
  • H04L2025/03522—Frequency domain

Definitions

• This invention relates to channel equalization and frame synchronization for Orthogonal Frequency Division Multiplexing (OFDM) data broadcast for non-mobile transmitters and receivers.
• OFDM Orthogonal Frequency Division Multiplexing
• OFDM Orthogonal Frequency Division Multiplexing
• OFDM transmission is an accurate modulation technique compared to other broadcasting techniques, there are two problems associated with it: one is frame synchronization and the other is channel equalization.
• OFDM data is broken up into individually modulated sections called frames. These frames are strung together in a continuous manner and transmitted over an OFDM channel. It is of utmost importance for the receiver to know precisely where the frames begin and end, in order to recover the valuable data. Finding these beginnings and endings of the frames is what is known as frame synchronization.
• Various techniques of performing frame synchronization exists, but they usually involves adding data flags for marking the beginning or the ending of the frames. This is done at the expanse of adding portions of data that slow down the speed of transmitting valuable data.
• OFDM channels sometimes distort the data they carry to the point that it is thoroughly unrecognizable, a scheme to correct any channel distortion is required.
• equalization techniques exist, but they often require the data correction to be done at the transmitter side so the state of the channel has to be known to the transmitter. Furthermore, these techniques do not apply to point to multipoint OFDM transmission since multiple paths are involved, each having its own different response.
• OFDM channel is used to designate the air path of the electromagnetic waves representing OFDM frames to be transmitted from a transmitting system to a receiving system.
• the transmitted OFDM information frames include reference frames, inserted at predetermined intervals, and these reference frames contain data known to the receiver. Upon receipt of these frames, the receiver uses them in order to collect data about the state of the OFDM channel. Having this information available, the receiver creates a correction or equalization vector, which is then multiplied with each of the information frames received in order to eliminate the undesirable effect of OFDM channel distortions.
• the end-result frames are verified for accuracy by multiplying the correction vector with a distorted reference frame and if the end-result is the accurate enough, the correction vector is considered to be reliable and may be used for correcting all subsequent information frames received. If the correction vector is inaccurate, another equalization vector may be created using other reference frames. If the vector is still inaccurate, then some pilot tones, consisting in very small samples of data known to the receiver and contained in the information frames, may further be used for updating the correction vector, for better results.
• ISI Inter-Symbol Interference
• a method of frame synchronization for OFDM channels wherein a frame synchronization starting point is to be located within an incoming data bit stream signal having repeated portions of data at predetermined intervals, wherein the method comprises the steps of: a) subtracting data collected from two points of said bit stream signal separated by a predetermined number of bits; b) detecting when said data is identical at said two points of said bit stream; c) obtaining said frame synchronization starting point from a result of step b).
• a method of channel equalization for point-to-multipoint OFDM channels is provided, wherein the method is performed at a receiver side of an OFDM channel and comprises the steps of: a) receiving an incoming data bit stream signal comprising interleaved sequences of reference and information frames, each of said reference frame containing data known to a receiver; b) detecting said reference frames in said incoming data bit stream signal of
• OFDM frames c) transforming said incoming stream of OFDM frames from a time domain to a frequency domain; d) creating a correction vector using said sequence of reference frames and data known to a receiver for correcting broadcast errors created by distortions caused by said OFDM channel; e) correcting each information frame from said sequences of information frames received at said receiver using said correction vector for countering a distortion effect of said OFDM channel on said frames.
• Figure 1 shows a preferred embodiment of the present invention related to the frame synchronization
• Figure 2 shows the cyclic prefix for two consecutive frames.
• Figure 3 represents a typical sequence of information frames preceded by a sequence of three reference frames used for frame synchronization at the receiver side;
• Figure 4 shows the application of the window or subtraction function on a sequence of incoming OFDM frames for roughly detecting the sequence of reference frames
• Figure 5 shows a preferred embodiment of the invention related to the fine reference frame search
• Figure 6 illustrates a flowchart of the initialization routine of receiver, according to a preferred embodiment of the present invention
• Figure 7 shows the information frame processing according to a preferred embodiment of the invention
• Figure 8 shows a detailed view of a preferred embodiment of the present invention related to the channel equalization
• a method for OFDM frame synchronization and channel equalization that allows accurate transmission of OFDM data without the need for the transmitter to know anything about the state of the channel.
• This method is especially suited for wireless non- mobile point-to-multipoint broad band data transmission, where multiple channels are involved, each having its own different response.
• a preferred embodiment of the present invention related to the frame synchronization is partially illustrated in Fig. 1. This figure shows the prior art technique of OFDM modulation at the transmitter side and the demodulation of data at the receiver side.
• Fig. 1 also comprises objects of the present invention which will be described in the following lines.
• a common string of bits 10 representing the valuable data to be transmitted from one point to another is fed into a digital string splitter 12 which splits the bit string 10 into separate words of bits 14, each having a pre-determined length. In our example, words of 6 bits are used.
• the result of this step is a complex number 18 of the form A+Bj which represents the processed 6 bit word 14.
• Data words are successively put into an array 19, thus forming a one-dimension array 19 of words 18 which has the dimension of one frame of information.
• an Inverse Fast Fourier Transform (IFFT) module 20 receives as an input a series comprising the real parts of the complex numbers 18 and the imaginary parts of that same complex numbers 18 and performs on that series an IFFT.
• the result is a Fourier transformation that creates an interleaved time series of data, so that the input series is converted from the frequency domain into the time domain.
• the output of the IFFT module 20 is a time series of complex numbers 22 of the form C+Dj wherein the first index represents the real parts C and the second the imaginary parts D.
• Each of these indices C and D represents the valuable portion of a data frame 30, as shown in Figure 1.
• a one-dimension complex array 23 contains all the complex numbers 22 which result from the IFFT operation. Since the time series represents 1024 consecutive words, the output of the IFFT module 20 is a pair of data frames 30, each representing 1024 words 14.
• the frame 30 has the form of an analog curve as shown in Fig. 1, but is in fact represented by a series of discrete points, each having its coordinates in a digital representation. This is common OFDM processing and is known in the prior art.
• Fig. 1 The next step of the method shown in Fig. 1 is a preferred embodiment of the present invention and relates to the frame synchronization.
• Frame synchronization means that the receiver has to detect a reference throughout the incoming data bit signal that indicates the beginning of the frames. Therefore, in a preferred embodiment of the invention, at the transmitter's side, each frame 30 comprising the valuable data to be transmitted enters a transmitter frame synchronization module 24 whose function is to add a reference in front of each frame in order to allow the receiver to detect the beginning the frames.
• Fig. 2 shows a better view of the frames comprising a cyclic prefix 25 composed of two identical segments 26.
• each frame 30 to be transmitted is transformed in such a way that a segment of data 26 of the frame is copied and appended to the beginning of the frame thus creating a longer frame 28, than the valuable data segment 30 that contains the useful data.
• This segment of valuable data 30, contains the useful data to be transmitted from a transmitter to one or more receivers.
• a small data segment 26 having preferably approximately 10% of the length of the valuable data portion 30 is copied and appended to the beginning of the portion 30, thus increasing the length of the frame by 10 % only.
• the process of creating a cyclic prefix 25 is performed on both the reference frames and the data frames and the result is that each transmitted frame begins and ends by the same segment of data 26.
• These identical segments 26 inside a frame 28 are used at the receiver side in order to detect the beginning of the frames, in a manner which is disclosed in detail further in this text.
• the transmitter preferably sends out at prescribed intervals sequences of three consecutive reference frames 34 between sequences of about 100 information frames 36. All reference frames contain identical data which is known to the receiver and these reference frames are used for frame synchronization purposes.
• An example of frame sequences sent out by the transmitter 8 is illustrated in Fig. 3.
• the receiver frame synchronization module 32 must perform two tasks, the first is to locate the sequence of reference frames 34 between two sequences of information frames 36 and the second is to precisely locate the first sample of data of the first reference frame 34 of that sequence of reference frames. Accurate detection of the beginning of the reference frames is needed in order to subsequently perform channel equalization.
• the reference frames 34 are frames of information known to the receiver 40 and are preferably sent by the transmitter 8 in sequences of three consecutive frames, at prescribed intervals among the information frames 36, as shown in Fig. 3. Other sequences comprising more or less reference frames may be used as well.
• the receiver 40 may use a frame synchronization module 38 in order to detect the location of the reference frames 34 by comparing sequences of data from the beginning and the end of the frame, separated by a length equivalent or bigger than a length of a frame.
• the principle, shown in Fig. 4, is to apply a subtraction between the two sections of data, one at the beginning of the reference frame 34 and the other at its end, which are supposed to be identical. In fact the two sequences are no longer identical, but only close to be, since the OFDM channel have distorted the data during the broadcast. This distortion is a difference in the form of the data between the transmitted data and the received data.
• the subtraction function is applied on all the range of the sequence of three reference frames, as shown in Fig. 4, and produces a minimum at the location of the first two reference frames from a sequence of three reference frames.
• a rough reference frame detection is performed.
• the presence of the sequence of three reference frames 34 is detected using an algorithm based on the following window function, or subtraction function:
• x is the sample index
• k is an arbitrary "window size”
• D is the data set
• N is the number of samples in one or more frames of data.
• the routine locates the location of the falling edge of the function W(x), it finds the location of the sequence of reference frame 34 only roughly.
• the purpose of this first searching routine is to locate the falling edge 39 of the subtraction function curve only roughly.
• Another routine is then employed to precisely find the location of the first sample of data of the reference frames 34. This is accomplished using a process similar to the one used at first.
• an ending segment 26 of each frame 22 is copied and appended at the beginning of that frame 22, thus creating a cyclic prefix 25 as shown in Fig. 2.
• the cyclic prefix 25 comprises the two identical segments of data 26, one being located at the beginning of the frame while the other is located at the end of the frame.
• the cyclic prefix 25 serves two functions: the first one is to protect frames from the impulse response spill-over generated by the previous frame, thus fortifying the system against ISI (Inter-Symbol Interference); the second and more crucial to frame synchronization is to provide a method of applying a variant of the W(x) function, call it w(x), such that w(x) encounters a minimum at the beginning of a reference frame 10.
• w(x) is then of the same form as W(x) but has different parameters: k is now the length in samples of the cyclic prefix, D is the input data set, N is the length in samples of one frame minus the extra part donated by the cyclic prefix and x remains the data index.
• This second routine is illustrated in Fig.
• the rough reference frame search and the fine reference frame search are only performed when the receiver is turned on, until the frame synchronization starting point is found. Once this is done, the processor may keep track of the beginning of each subsequent frame, since all frames have a predetermined length, so these steps may be skipped after the initialization of the receiver. However, when distortions reach a point where keeping track of the beginning of the frames is no longer possible, the process of resetting the receiver and redoing the search for the beginning of the reference frames may be restarted.
• a time domain correlation algorithm may be applied to the reference data in order to correct the frequency offset, which is a data distortion phenomena caused by the small variations of the local oscillators from the transmitter and receiver's sides.
• the principle is to perform a correlation between the first and the second reference frame in the time domain, by using a correlation coefficient ⁇ in the time domain.
• ⁇ is defined as:
• Nfft is the size of the FFT; the asterisk denotes the complex conjugate; x2 is the starting point obtained from the fine reference frame search; In a variant of this point of the technique, it may be useful to average two value of ⁇ in order to obtain a more accurate value.
• a first value of ⁇ is obtained by computing the first and second reference frame while the second value is obtained computing the second and third reference frames from the sequence of three reference frames. Such a value of ⁇ better represents the frequency offset correction for an average of three consecutive reference frames received over a longer period of time.
• Fig. 6 may be considered a continuation of Fig. 1 and mainly shows a flowchart of the receiver's channel equalization module 50.
• frame detection is performed using the window function described above and its variant.
• the next step after finding the beginning of the reference frame and hence all frames since the CPU can keep track of the input index, is to transform the data into the frequency domain. This is accomplished using an FFT algorithm.
• the radix-4 method may be used and the processing is done in real-time.
• the preferred embodiment of the present invention is an equalization routine for OFDM channels which works on the frequency domain data.
• An equalization vector employed for equalizing the incoming data bit signal may be updated at each incidence of the three reference frames which occurs every 100 frames approximately. It was mentioned earlier that the incoming data bit signal comprises reference frames 34 that are data known to the receiver and a sample 45 of a reference frame data is hardwired at the receiver's side into a memory 52, such as an EEPROM, onboard the receiver's circuitry.
• Fig. 8 shows a detailed receiver's channel equalization flowchart. After the reference frames 34 have been located by the frame detection routines, the first step is to turn them into frequency domain, using an FFT module 54.
• each of the two first reference frames 34 received over the OFDM channel is divided by the accurate reference frame sample 45 contained in the memory 52 and the two results are the inverses of the frequency response of the OFDM channel.
• the results are averaged and the result is a more accurate correction or equalization vector which represents the average distortion caused by the OFDM channel on two consecutive frames.
• the correlation coefficient is also involved in the mathematical operation in order to counter the problem of the frequency offset, described in detail earlier in this application.
• the formula used is:
• Veq is the averaged equalization vector
• Veql is the first equalization vector
• Veq2 is the second equalization vector
• i is the imaginary unity
• ⁇ is the correlation coefficient
• This operation results in a more accurate correction vector representing the inverse of the frequency response of the communication channel.
• Other smoothing algorithms may be applied to this vector in order to filter spurious spikes and noise.
• the correction vector is then multiplied, component by component to the samples extracted from the third reference frame received inside the sequence of three reference frames, thereby equalizing it and removing the distortion effect of the channel.
• This corrected third reference frame should then be equal to the accurate reference frame 45 contained in the receiver's memory 52.
• the verification may be performed more than once before the routine continues, and each time the operation is done and the result is not accurate enough, the FFT start point of the third reference frame in the sequence may be shifted back one sample of data.
• each subsequent information frame 36 is multiplied by the same equalization vector according to the following algorithm:
• Ieq is the corrected data frame
• I is the distorted data frame
• Fig. 6 shows a general functioning flowchart of the receiver initialization routine that is performed each time the receiver is turned on. The purpose of this routine is to locate a best point in time for starting reading the incoming data and from that point, to continue keeping track of the starting points of each incoming frame of data. As it can be seen in Fig.
• the reference frames are correlated into the time domain for correcting the effect of the frequency offset. Then, an FFT is performed on the reference frames and a first equalization vector is generated according to the method shown in Fig. 8. This vector goes through a plurality of verification involving comparison of the result with the accurate reference frame 45, until a best starting point is found for the subsequent reading of incoming data.
• Fig. 7 shows the information frame processing, once the accurate equalization vector has been generated. All incoming information frames go through a process where the cyclic prefix 25 is removed and are then correlated in the time domain for removing the effect of the frequency offset. Then, they all go through an FFT operator and are multiplied by the equalization vector in the manner better explained earlier in this text, for removing the distortion caused on them by the OFDM channel. Finally, the frames are de-mapped and the digital data is output to the receiver data processing unit. In another preferred embodiment of the present invention, further data verification may be performed at the receiver's side in order to assure that the data is restored in an accurate form. A verification of the condition of the restored information frames may be performed using very small samples of information contained in each information frame 36 and called pilot tones.
• pilot tones are very small sequences of data containing information known to the receiver, just like the reference frames. Therefore, after each information frame is corrected using the correcting vector, these pilot tones can be verified and if they begin to drift from their expected values, a feedback signal may be sent in order to update the correction vector using the drift-values of the pilot tones.
• the equalization vector is then only updated at each incidence of the reference frames, which occurs about once each 100 data frames, and the data frames are updated using the equalization vector, as described above.
• the updating of the equalization vector may be performed even more rarely, for example at each third incidence of the sequence of reference frames.
• the rough reference frame search and the fine reference frame search do not need to be performed anymore, since the exact location of the reference frames is known. This saves time and increases the performance of the receiver.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Power Engineering (AREA)
• Synchronisation In Digital Transmission Systems (AREA)
• Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)

Priority Applications (12)

Applications Claiming Priority (2)

Publications (3)

Family

ID=21856418

Family Applications (1)

Country Status (17)

Cited By (12)

Families Citing this family (39)

Citations (3)

Family Cites Families (23)

• 1998
  • 1998-02-26 US US09/030,865 patent/US6151295A/en not_active Expired - Lifetime
• 1999
  • 1999-02-25 ID IDW20001915A patent/ID26957A/id unknown
  • 1999-02-25 PL PL99343241A patent/PL343241A1/xx unknown
  • 1999-02-25 CN CNB998033073A patent/CN1248474C/zh not_active Expired - Fee Related
  • 1999-02-25 JP JP2000533973A patent/JP2002505551A/ja active Pending
  • 1999-02-25 EP EP99936153A patent/EP1078502A2/en not_active Withdrawn
  • 1999-02-25 WO PCT/CA1999/000175 patent/WO1999044326A2/en active IP Right Grant
  • 1999-02-25 EP EP03015533A patent/EP1361717A3/en not_active Withdrawn
  • 1999-02-25 UA UA2000095454A patent/UA56302C2/uk unknown
  • 1999-02-25 AU AU32425/99A patent/AU3242599A/en not_active Abandoned
  • 1999-02-25 BR BR9908302-7A patent/BR9908302A/pt not_active IP Right Cessation
  • 1999-02-25 EA EA200000882A patent/EA003198B1/ru not_active IP Right Cessation
  • 1999-02-25 IL IL13789499A patent/IL137894A0/xx not_active IP Right Cessation
  • 1999-02-25 CA CA002320058A patent/CA2320058C/en not_active Expired - Fee Related
  • 1999-02-25 KR KR1020007009401A patent/KR100729260B1/ko not_active IP Right Cessation
• 2000
  • 2000-08-24 NO NO20004252A patent/NO20004252L/no not_active Application Discontinuation
  • 2000-09-13 BG BG104762A patent/BG104762A/bg unknown
• 2001
  • 2001-06-21 HK HK01104348A patent/HK1034009A1/xx not_active IP Right Cessation

Patent Citations (3)

Non-Patent Citations (3)

Also Published As

Similar Documents

Legal Events